Abstract

Simple Reduction and Beyond Traditional and New Wave models of reduction in science have not lacked for ambition. Philosophers have presented single models to account for the full range of interesting intertheoretic relations, for scientific progress, and for the unity of science (Nagel, 1961; Oppenheim and Putnam, 1958). Early critics attacked the logical empiricists' proposals about the character of intertheoretic connections (Feyerabend, 1962; Kuhn, 1970). New Wave reductionists have similarly argued that various intertheoretic relations fall at different points on a continuum of goodness-of-intertheoretic-mapping. Still, whatever their differences with the logical empiricists, New Wave reductionists have retained traditional aspirations for a single, comprehensive model of reduction that will make sense of the wide range of intertheoretic relations, of progressive scientific change, and of how the various sciences hang together (Hooker, 1981; Churchland and Churchland, 1990; Bickle, 1998). Both logical empiricists and their New Wave successors proffer such unified, multi-purpose models. Regardless of the field, multi-purpose tools typically sacrifice precision for versatility. Recent analyses of mechanistic explanation have helped to reveal that these models of scientific reduction are no exceptions to this rule, and the cost of sacrificing precision is one that the mechanists are unwilling to pay. Traditional and New Wave reductionists manifest allegiances (1) to the (virtually exclusive) analysis of theories and intertheoretic relations and (2) to conceptions of explanatory levels in science rooted in considerations pertaining to the size of and the mereological relations between the sciences' objects of study. (Discussions of both explanatory levels and mereological relations follow in subsequent sections.) By contrast, advocates of mechanistic analysis offer detailed accounts of particular systems' functioning that survey their components, their operations, and the larger systems to which they contribute (Bechtel and Richardson, 1993; Glennan, 1996; Machamer, Darden and Craver, 2000; Craver, 2001; Bechtel, 2006)..